In large software runtime environments, such as, server environments, it is often difficult to diagnose problems in a system because any number of seemingly unrelated factors may interact and result in performance issues or system failure. Information useful in diagnosing system problems, such as logs, performance measurements, and so forth, may also be scattered in a variety of places. Additionally, less experienced users may not know which types of information are relevant for a particular system problem. Diagnosing system problems is further complicated by the fact that information necessary for diagnosis may need to be viewed using multiple viewers that are unlikely to be synchronized. Hence, a big picture view of what was happening in the system when a particular problem occurred will not be available.